[PATH 1/1] Kexec jump - v2 - kexec jump

From: Huang, Ying
Date: Sun Jul 15 2007 - 03:15:13 EST

This patch implement the functionality of jumping from kexeced kernel
to original kernel.

A new reboot command named LINUX_REBOOT_CMD_KJUMP is defined to
trigger the jumping to (executing) the new kernel or jumping back to
the original kernel.

To support jumping back from kexeced kernel, before executing the new
kernel, the devices are put into quiescent state (to be fully
implemented), and the state of devices and CPU is saved. After jumping
back from kexeced kernel, the state of devices and CPU are restored
accordingly. The devices/CPU state save/restore code of software
suspend is called to implement corresponding function.

Signed-off-by: Huang Ying <ying.huang@xxxxxxxxx>

---

arch/i386/Kconfig | 7 +++
arch/i386/kernel/Makefile | 1
arch/i386/kernel/kexec_jump.S | 74 +++++++++++++++++++++++++++++++++++++
arch/i386/kernel/machine_kexec.c | 59 +++++++++++++++++++++++++++++
include/asm-i386/kexec.h | 4 ++
include/linux/kexec.h | 17 ++++++++
include/linux/reboot.h | 1
kernel/kexec.c | 78 +++++++++++++++++++++++++++++++++++++++
kernel/ksysfs.c | 12 ++++++
kernel/sys.c | 8 ++++
10 files changed, 261 insertions(+)

Index: linux-2.6/arch/i386/kernel/Makefile
===================================================================
--- linux-2.6.orig/arch/i386/kernel/Makefile 2007-07-14 14:52:45.000000000 +0800
+++ linux-2.6/arch/i386/kernel/Makefile 2007-07-14 17:50:50.000000000 +0800
@@ -27,6 +27,7 @@
obj-$(CONFIG_X86_REBOOTFIXUPS) += reboot_fixups.o
obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o crash.o
obj-$(CONFIG_CRASH_DUMP) += crash_dump.o
+obj-$(CONFIG_KEXEC_JUMP) += kexec_jump.o
obj-$(CONFIG_X86_NUMAQ) += numaq.o
obj-$(CONFIG_X86_SUMMIT_NUMA) += summit.o
obj-$(CONFIG_KPROBES) += kprobes.o
Index: linux-2.6/arch/i386/kernel/machine_kexec.c
===================================================================
--- linux-2.6.orig/arch/i386/kernel/machine_kexec.c 2007-07-14 14:52:45.000000000 +0800
+++ linux-2.6/arch/i386/kernel/machine_kexec.c 2007-07-15 11:22:23.000000000 +0800
@@ -10,6 +10,7 @@
#include <linux/kexec.h>
#include <linux/delay.h>
#include <linux/init.h>
+#include <linux/highmem.h>
#include <asm/pgtable.h>
#include <asm/pgalloc.h>
#include <asm/tlbflush.h>
@@ -169,3 +170,61 @@
return 0;
}
early_param("crashkernel", parse_crashkernel);
+
+#ifdef CONFIG_KEXEC_JUMP
+int machine_kexec_jump(void)
+{
+ asmlinkage int (*real_jump)(void *buf);
+ unsigned long jump_buf_pfn;
+ void *jump_buf;
+
+ if (!(kexec_flags & KEXEC_FLAGS_IS_KEXECED_KERNEL) &&
+ !kexec_crash_image)
+ return -EINVAL;
+ jump_buf_pfn = kexec_get_jump_buf_pfn(0);
+ if (!jump_buf_pfn)
+ return -EINVAL;
+ jump_buf = kmap_atomic_pfn(jump_buf_pfn, KM_PTE0);
+ memcpy(jump_buf + PAGE_SIZE/2, machine_kexec_real_jump, PAGE_SIZE/2);
+ real_jump = jump_buf + PAGE_SIZE/2;
+
+ if (!real_jump(jump_buf))
+ machine_kexec(kexec_crash_image);
+ kunmap_atomic(jump_buf, KM_PTE0);
+ return 0;
+}
+
+static unsigned long kexec_backup_addr = ~0UL;
+
+/* kexec_backup= specifies the location of backuped 0~640k memory of
+ * crashed kernel.
+ */
+static int __init parse_kexec_backup(char *arg)
+{
+ if (!arg)
+ return -EINVAL;
+
+ kexec_backup_addr = memparse(arg, &arg);
+ return 0;
+}
+early_param("kexec_backup", parse_kexec_backup);
+
+void kexec_restore_backup(void)
+{
+ void *vaddr;
+ void *vaddr_backup;
+ unsigned long paddr;
+
+ if (kexec_backup_addr == ~0UL)
+ return;
+
+ for (paddr = 0; paddr < 640 * 1024; paddr += PAGE_SIZE) {
+ vaddr = kmap_atomic_pfn(paddr >> PAGE_SHIFT, KM_PTE0);
+ vaddr_backup = kmap_atomic_pfn((paddr+kexec_backup_addr) >> PAGE_SHIFT,
+ KM_PTE1);
+ memcpy(vaddr, vaddr_backup, PAGE_SIZE);
+ kunmap_atomic(vaddr, KM_PTE0);
+ kunmap_atomic(vaddr_backup, KM_PTE1);
+ }
+}
+#endif /* CONFIG_KEXEC_JUMP */
Index: linux-2.6/include/asm-i386/kexec.h
===================================================================
--- linux-2.6.orig/include/asm-i386/kexec.h 2007-07-14 14:52:45.000000000 +0800
+++ linux-2.6/include/asm-i386/kexec.h 2007-07-14 21:21:16.000000000 +0800
@@ -94,6 +94,10 @@
unsigned long start_address,
unsigned int has_pae) ATTRIB_NORET;

+#ifdef CONFIG_KEXEC_JUMP
+extern asmlinkage int machine_kexec_real_jump(void *buf);
+#endif
+
#endif /* __ASSEMBLY__ */

#endif /* _I386_KEXEC_H */
Index: linux-2.6/include/linux/kexec.h
===================================================================
--- linux-2.6.orig/include/linux/kexec.h 2007-07-14 14:52:45.000000000 +0800
+++ linux-2.6/include/linux/kexec.h 2007-07-15 11:22:57.000000000 +0800
@@ -1,6 +1,11 @@
#ifndef LINUX_KEXEC_H
#define LINUX_KEXEC_H

+#ifdef CONFIG_KEXEC_JUMP
+#define KEXEC_FLAGS_IS_KEXECED_KERNEL 0x1
+#endif /* CONFIG_KEXEC_JUMP */
+
+#ifndef __ASSEMBLY__
#ifdef CONFIG_KEXEC
#include <linux/types.h>
#include <linux/list.h>
@@ -161,4 +166,16 @@
static inline void crash_kexec(struct pt_regs *regs) { }
static inline int kexec_should_crash(struct task_struct *p) { return 0; }
#endif /* CONFIG_KEXEC */
+
+#ifdef CONFIG_KEXEC_JUMP
+extern int kexec_flags;
+extern int machine_kexec_jump(void);
+extern int kexec_jump(void);
+extern unsigned long kexec_get_jump_buf_pfn(int alloc);
+extern void kexec_restore_backup(void);
+#else
+static inline int kexec_jump(void) { return 0; }
+#endif /* CONFIG_KEXEC_JUMP */
+#endif /* !__ASSEMBLY__ */
+
#endif /* LINUX_KEXEC_H */
Index: linux-2.6/kernel/kexec.c
===================================================================
--- linux-2.6.orig/kernel/kexec.c 2007-07-14 14:52:45.000000000 +0800
+++ linux-2.6/kernel/kexec.c 2007-07-15 12:57:25.000000000 +0800
@@ -22,6 +22,9 @@
#include <linux/hardirq.h>
#include <linux/elf.h>
#include <linux/elfcore.h>
+#include <linux/suspend.h>
+#include <linux/pm.h>
+#include <linux/cpu.h>

#include <asm/page.h>
#include <asm/uaccess.h>
@@ -1135,3 +1138,78 @@
return 0;
}
module_init(crash_notes_memory_init)
+
+#ifdef CONFIG_KEXEC_JUMP
+int kexec_flags;
+
+static unsigned long kexec_jump_buf_pfn;
+
+static int __init parse_kexec_jump_buf_pfn(char *arg)
+{
+ if (!arg)
+ return -EINVAL;
+
+ kexec_jump_buf_pfn = memparse(arg, &arg);
+ kexec_flags |= KEXEC_FLAGS_IS_KEXECED_KERNEL;
+ return 0;
+}
+early_param("kexec_jump_buf_pfn", parse_kexec_jump_buf_pfn);
+
+unsigned long kexec_get_jump_buf_pfn(int alloc)
+{
+ struct page *jump_buf_page;
+
+ if (!kexec_jump_buf_pfn && alloc) {
+ jump_buf_page = alloc_page(GFP_KERNEL);
+ kexec_jump_buf_pfn = page_to_pfn(jump_buf_page);
+ }
+ return kexec_jump_buf_pfn;
+}
+
+int kexec_jump(void)
+{
+ int error;
+
+ if (!(kexec_flags & KEXEC_FLAGS_IS_KEXECED_KERNEL) &&
+ !kexec_crash_image)
+ return -EINVAL;
+ error = device_suspend(PMSG_FREEZE);
+ if (error)
+ goto Resume_devices;
+ error = disable_nonboot_cpus();
+ if (error)
+ goto Enable_cpus;
+ local_irq_disable();
+ /* At this point, device_suspend() has been called, but *not*
+ * device_power_down(). We *must* device_power_down() now.
+ * Otherwise, drivers for some devices (e.g. interrupt controllers)
+ * become desynchronized with the actual state of the hardware
+ * at resume time, and evil weirdness ensues.
+ */
+ error = device_power_down(PMSG_FREEZE);
+ if (error)
+ goto Enable_irqs;
+
+ if (kexec_flags & KEXEC_FLAGS_IS_KEXECED_KERNEL)
+ kexec_restore_backup();
+ save_processor_state();
+ error = machine_kexec_jump();
+ restore_processor_state();
+
+ /* NOTE: device_power_up() is just a resume() for devices
+ * that suspended with irqs off ... no overall powerup.
+ */
+ device_power_up();
+ Enable_irqs:
+ local_irq_enable();
+ Enable_cpus:
+ enable_nonboot_cpus();
+ Resume_devices:
+ device_resume();
+ if (!error) {
+ kimage_free(kexec_crash_image);
+ kexec_crash_image = NULL;
+ }
+ return error;
+}
+#endif /* CONFIG_KEXEC_JUMP */
Index: linux-2.6/kernel/ksysfs.c
===================================================================
--- linux-2.6.orig/kernel/ksysfs.c 2007-07-14 14:52:45.000000000 +0800
+++ linux-2.6/kernel/ksysfs.c 2007-07-14 17:55:46.000000000 +0800
@@ -60,6 +60,15 @@
return sprintf(page, "%d\n", !!kexec_crash_image);
}
KERNEL_ATTR_RO(kexec_crash_loaded);
+
+#ifdef CONFIG_KEXEC_JUMP
+static ssize_t kexec_jump_buf_pfn_show(struct kset *kset, char *page)
+{
+ return sprintf(page, "0x%lx\n", kexec_get_jump_buf_pfn(1));
+}
+
+KERNEL_ATTR_RO(kexec_jump_buf_pfn);
+#endif
#endif /* CONFIG_KEXEC */

decl_subsys(kernel, NULL, NULL);
@@ -73,6 +82,9 @@
#ifdef CONFIG_KEXEC
&kexec_loaded_attr.attr,
&kexec_crash_loaded_attr.attr,
+#ifdef CONFIG_KEXEC_JUMP
+ &kexec_jump_buf_pfn_attr.attr,
+#endif
#endif
NULL
};
Index: linux-2.6/arch/i386/kernel/kexec_jump.S
===================================================================
--- /dev/null 1970-01-01 00:00:00.000000000 +0000
+++ linux-2.6/arch/i386/kernel/kexec_jump.S 2007-07-15 11:26:51.000000000 +0800
@@ -0,0 +1,74 @@
+/*
+ * kexec_jump.S - Jump between normal kernel and kexeced kernel
+ * Copyright (C) 2007 Huang Ying <ying.huang@xxxxxxxxx>
+ *
+ * This source code is licensed under the GNU General Public License,
+ * Version 2. See the file COPYING for more details.
+ */
+
+#include <linux/linkage.h>
+#include <linux/kexec.h>
+#include <asm/page.h>
+#include <asm/kexec.h>
+
+/*
+ * Must be relocatable PIC code callable as a C function
+ */
+#define HALF_PAGE_ALIGNED (1 << (PAGE_SHIFT-1))
+
+#define EBX 0x0
+#define ESI 0x4
+#define EDI 0x8
+#define EBP 0xc
+#define ESP 0x10
+#define CR0 0x14
+#define CR3 0x18
+#define CR4 0x1c
+#define FLAG 0x20
+#define RET 0x24
+
+ .text
+ .align HALF_PAGE_ALIGNED
+ .globl machine_kexec_real_jump
+machine_kexec_real_jump:
+ movl 4(%esp), %edx
+ movl kexec_flags, %ecx
+ andl $KEXEC_FLAGS_IS_KEXECED_KERNEL, %ecx
+ jnz 1f
+ movl %ebx, EBX(%edx)
+ movl %esi, ESI(%edx)
+ movl %edi, EDI(%edx)
+ movl %ebp, EBP(%edx)
+ movl %esp, ESP(%edx)
+ movl %cr0, %eax
+ movl %eax, CR0(%edx)
+ movl %cr3, %eax
+ movl %eax, CR3(%edx)
+ movl %cr4, %eax
+ movl %eax, CR4(%edx)
+ pushf
+ popl %eax
+ movl %eax, FLAG(%edx)
+ movl (%esp), %eax
+ movl %eax, RET(%edx)
+ mov $0, %eax
+ ret
+1:
+ movl EBX(%edx), %ebx
+ movl ESI(%edx), %esi
+ movl EDI(%edx), %edi
+ movl EBP(%edx), %ebp
+ movl FLAG(%edx), %eax
+ pushl %eax
+ popf
+ movl ESP(%edx), %esp
+ movl CR4(%edx), %eax
+ movl %eax, %cr4
+ movl CR3(%edx), %eax
+ movl %eax, %cr3
+ movl CR0(%edx), %eax
+ movl %eax, %cr0
+ movl RET(%edx), %eax
+ movl %eax, (%esp)
+ mov $1, %eax
+ ret
Index: linux-2.6/kernel/sys.c
===================================================================
--- linux-2.6.orig/kernel/sys.c 2007-07-14 16:45:21.000000000 +0800
+++ linux-2.6/kernel/sys.c 2007-07-14 17:39:47.000000000 +0800
@@ -940,6 +940,14 @@
unlock_kernel();
return -EINVAL;

+ case LINUX_REBOOT_CMD_KJUMP:
+ {
+ int ret;
+ ret = kexec_jump();
+ unlock_kernel();
+ return ret;
+ }
+
#ifdef CONFIG_SOFTWARE_SUSPEND
case LINUX_REBOOT_CMD_SW_SUSPEND:
{
Index: linux-2.6/include/linux/reboot.h
===================================================================
--- linux-2.6.orig/include/linux/reboot.h 2007-07-14 17:29:58.000000000 +0800
+++ linux-2.6/include/linux/reboot.h 2007-07-14 20:43:04.000000000 +0800
@@ -33,6 +33,7 @@
#define LINUX_REBOOT_CMD_RESTART2 0xA1B2C3D4
#define LINUX_REBOOT_CMD_SW_SUSPEND 0xD000FCE2
#define LINUX_REBOOT_CMD_KEXEC 0x45584543
+#define LINUX_REBOOT_CMD_KJUMP 0x3928A5FD


#ifdef __KERNEL__
Index: linux-2.6/arch/i386/Kconfig
===================================================================
--- linux-2.6.orig/arch/i386/Kconfig 2007-07-14 17:52:02.000000000 +0800
+++ linux-2.6/arch/i386/Kconfig 2007-07-14 19:08:37.000000000 +0800
@@ -813,6 +813,13 @@
PHYSICAL_START.
For more details see Documentation/kdump/kdump.txt

+config KEXEC_JUMP
+ bool "kexec jump (EXPERIMENTAL)"
+ depends on EXPERIMENTAL
+ depends on PM && X86_32 && KEXEC
+ ---help---
+ Jump back from kexeced kernel to orignal kernel.
+
config PHYSICAL_START
hex "Physical address where the kernel is loaded" if (EMBEDDED || CRASH_DUMP)
default "0x100000"
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/